| Presentation # | 1 |
| Session: | Spoken Language Understanding |
| Session Time: | Wednesday, December 19, 10:00 - 12:00 |
| Presentation Time: | Wednesday, December 19, 10:00 - 12:00 |
| Presentation: |
Poster
|
| Topic: |
Spoken language understanding: |
| Paper Title: |
LOW-RESOURCE CONTEXTUAL TOPIC IDENTIFICATION ON SPEECH |
| Authors: |
Chunxi Liu; Johns Hopkins University | | |
| | Matthew Wiesner; Johns Hopkins University | | |
| | Shinji Watanabe; Johns Hopkins University | | |
| | Craig Harman; Johns Hopkins University | | |
| | Jan Trmal; Johns Hopkins University | | |
| | Najim Dehak; Johns Hopkins University | | |
| | Sanjeev Khudanpur; Johns Hopkins University | | |
| Abstract: |
In topic identification (topic ID) on real-world unstructured audio, an audio instance of variable topic shifts is first broken into sequential segments, and each segment is independently classified. We first present a general purpose method for topic ID on spoken segments in low-resource languages, using a cascade of universal acoustic modeling, translation lexicons to English, and English-language topic classification. Next, instead of classifying each segment independently, we demonstrate that exploring the contextual dependencies across sequential segments can provide large improvements. In particular, we propose an attention-based contextual model which is able to leverage the contexts in a selective manner. We test both our contextual and non-contextual models on four LORELEI languages, and on all but one our attention-based contextual model significantly outperforms the context-independent models. |